In familial cases of neurodegenerative diseases, mutations are found that usually alter protein coding sequences. In contrast, although there are genes that act as risk factors for sporadic disease, these are not usually associated with protein coding changes. This leads to the concept that common genetic variants alter mRNA expression as a quantitative trait. Within a given human population, normal genetic variation at many different loci will show association with level of mRNA expression. However, gene expression is subject to regulation and multiple levels between DNA and mature, stable, mRNA. Genomewide approaches are starting to be applied to such regulatory steps, exemplified by high throughput approaches to epigenetic modifications of DNA and to microRNA (miRNA). [unreadable] Over the past year, we have developed a new project where our objective is to build up a framework for how genotype and expression are related within and between different regions of the human brain. We have measured mRNA and miRNA expression patterns in a moderately sized series of control human brains. We included four different brain regions to begin to capture the complexity of gene expression across this complex organ. Both independent measures mRNA and miRNA showed distinct profiles across different brain regions. We are currently analyzing the relationship(s) between miRNA and mRNA expression levels and between genotype and RNA. In this way, we are building up a more detailed picture of the relationship between genotype and expression in that we include some of the modulatory steps in between DNA and RNA. [unreadable] Additionally, we have also begun to explore deep resequencing of RNA from brain. In this technique, we isolate either small RNA species (including miRNA) or polyadenylated full length mRNA and sequence fragments using a Solexa/Illumina Genome Analyzer II, which can generate up to 2 billion base pair reads per run in fragments of 35-40 nucleotides each. This has several advantages over conventional arrays in that we obtain sequence information directly, along with counts of frequency. We also have the potential to discover novel expressed sequences. To provide feasibility data, we took a small set of brain samples that we had used in the above experiments and generated sequence data for purified small RNA (20-40 nucleotides) and for polyA RNA. We also used frontal cortex samples from non-human primates to compare gene expression across species, which would be much more difficult with conventional techniques. Analysis of this data is ongoing.